Conventionally, there is a technology of adding marker images to a projection image for specifying the four corners of the projection image. A projection device projects such a projection image onto a projection object such as a whiteboard. A photographing device photographs the projection image. A detecting device detects, in the photographed image, the areas in which the marker images are projected. Accordingly, projection distortions in the photographed image can be corrected before displaying the image with a remote display device. Furthermore, coordinates that are pointed out on the projection object can be converted into the coordinates of the actual image, and the converted coordinates can be recorded.
FIG. 18 illustrates an example of converting coordinates that are pointed out on the projection object into the coordinates of the actual image before recording the converted coordinates. An auxiliary marker image 90E is projected by a projection device in addition to marker images 90A through 90D at the four corners of a projection image. Accordingly, erroneous detections of the marker images 90A through 90D are prevented, and the precision in converting coordinates is improved (see, for example, patent document 1).    Patent Document 1: Japanese Laid-Open Patent Application No. 2008-225553
However, the above-described conventional technology has the following problem. Suppose that the projection device, the projection object, and the photographing device have been positioned, and the projection image area in which the projection image is projected has been detected based on a photographed image. However, if the relative positions of the projection device, the projection object, and the photographing device change due to some reason such as vibration, the detected projection image area may become irrelevant. Accordingly, an incorrect projection image area is recognized, unless the projection image area is detected once again by user operation (after the vibration).